IoT data is inherently noisy. There are often cases of invalid and missing values. This requires constant vigilance to identify and correct data issues when they occur. The correction could then be handled in the transformation of the raw data or in the software and design of the device.

Either way, the faster an issue is detected, the quicker it can be resolved. For IoT data, consider bad data as lost money that can rarely be recovered. Minimize the loss by identifying and correcting issues quickly.

Dashboards can also be created for this purpose by following the same process we introduced in this chapter. Think about what you want to watch out for and set up an alert view to identify it for you.